Supervisors: Dr. Ali Meghdari, Dr. Minoo Alemi
Simultaneous Localization And Mapping (SLAM) is regarded as one of the most significant advances in robotics in recent decades. It enables robots to map their surrounding environments without the need for external equipment and simultaneously determine their position on this map. In the present study, this technique was implemented on Arash social robot to enable it to estimate its position in the environment with decent accuracy and potentially be able to automatically move towards the targets determined on the map using a path planning algorithm. At first, Arash robot mechanically was redesigned for this technique, and then the hardware and software needed to high-level and low-level control of the robot were developed. By conducting the literature review of SLAM techniques, the FastSLAM algorithm was found to be suitable for employing on Arash robot, and it was implemented using Robotic Operating System (ROS). Then, to improve the quality of the SLAM technique in crowded and dynamic environments, a person detection and tracking module were implemented on the Arash robot. In the experiments, it was observed that this module improves the quality of mapping and pose estimation of the robot in a crowded environment by filtering the data of moving people from the laser rangefinder sensor and publishing these data to the FastSLAM algorithm.